Shenghong Li

Wuhan University

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

Jianhua Li

Shanghai Jiao Tong University

Shilin Wang

Zhongnan Hospital of Wuhan University

Hao Ge

Nanjing University of Science and Technology

Chong Di

Qilu University of Technology

Chenglin Zhao

Beijing University of Posts and Telecommunications

Kangkang Wang

Xinjiang University

Ying Guo

Research Institute of Highway

Wen Jiang

Northwestern Polytechnical University

Gaolei Li

Shanghai Jiao Tong University

Haonan Guo

Liaoning Technical University

Cooperative Institutions

Shanghai Jiao Tong University

136

New Horizons Diagnostics (United States)

Chinese Academy of Sciences

Zhejiang University

Tsinghua University

Huazhong University of Science and Technology

Peking University

Beijing University of Posts and Telecommunications

Central South University

Harbin Institute of Technology

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

Stochastic stability and the moment Lyapunov exponent for a gyro-pendulum system driven by a bounded noise

Mechanical sciences (2023)

Shenghong Li Junting Lv

Abstract. The stochastic stability of a gyro-pendulum system parametrically excited by a real noise is investigated by the moment Lyapunov exponent in the paper. Using the spherical polar and non-singular linear stochastic transformations and combining these with Khasminskii's method, the diffusion process and the eigenvalue problem of the moment Lyapunov exponent are obtained. Then, applying the perturbation method and Fourier cosine series expansion, we derive an infinite-order matrix whose leading eigenvalue is the second-order expansion g2(p) of the moment Lyapunov exponent. Thus, an infinite sequence for g2(p) is constructed, and its convergence is numerically verified. Finally, the influences of the system and noise parameters on stochastic stability are given such that the stochastic stability is strengthened with the increased drift coefficient and the diffusion coefficient has the opposite effect; among the system parameters, only the increase in k and A0 strengthens moment stability.

10.5194/ms-14-545-2023

Cite

Citations (0)

Identification of influential nodes in social networks with community structure based on label propagation

Neurocomputing (2016)

Yuxin Zhao Shenghong Li Feng Jin

Identification

Maximization

Social network (sociolinguistics)

Evolving networks

Dynamic network analysis

Network Structure

10.1016/j.neucom.2015.11.125

Cite

Citations (96)

Adaptive continuous action‐set learning automata scheme

Electronics Letters (2018)

Ying Guo Shenghong Li Bo Fan

Learning automata (LA) has been widely encountered in signal processing for images, speeches, videos and so on. One special type of LA, i.e. the CALA (continuous action‐set LA) issue is handled. It first fills the existing gap aimed at the classic Beigy's algorithm, based on which a novel adaptive CALA (ACALA) algorithm is proposed. The proposed ACALA algorithm includes a sampling process and an iterating process, where the learning parameters are adaptively adjusted, not only between the two stages but also inside the core stage. Experiments with regard to noisy function optimisation reveal its outperformance, in accordance with the general evaluation system built. Specifically, the final result of the proposed algorithm could always judge as the global optimum even in noise‐added scenarios, regardless of the initial parameters.

Learning Automata

10.1049/el.2017.4014

Cite

Citations (2)

Linear transceiver design for full-duplex multi-user MIMO system

Shenghong Li Ross Murch Vincent K. N. Lau

We focus on a full-duplex multi-user MIMO system where the base station (BS) serves multiple uplink and downlink users simultaneously. Both the BS and the mobile stations (MSs) are equipped with multiple antennas. The performance of the system is limited by the self-interference at the base station and the interference caused by the uplink users on the downlink users. To address this issue, we propose to jointly design the beamformers of the BS and MSs. An optimization problem is formulated to maximize the weighted sum data rate subject to maximum power constraints. Although the problem is non-convex, it can be solved via iterative minimization of weighted sum mean square error (MSE), and a stationary point of the problem can be obtained. Strategies to choose uplink and downlink users and to determine initial point for the optimization problem are also proposed. Simulation results show that the weighted sum data rate achieved by full-duplex system is substantially higher than that achieved by baseline half-duplex systems.

Duplex (building)

Transmitter power output

Transceiver

Minification

10.1109/icc.2014.6884100

Cite

Citations (40)

Influence Maximization on Dynamic Social Networks with Conjugate Learning Automata

arXiv (Cornell University) (2020)

Fangqi Li Chong Di Shenghong Li

Selecting the optimal subset from all vertices as seeds to maximize the influence in a social network has been a task of interest. Various methods have been proposed to select the optimal vertices in a static network, however, they are challenged by the dynamics, i.e. the time-dependent variation of the social network structure. Such dynamics hinder the paradigm for static networks and leaves a seemingly unbridgeable gap between algorithms of influence maximization on static networks and those on dynamic ones. In this paper, we extend our previous work and demonstrate that conjugate learning automata (an elementary variant of reinforcement learning) that have been successfully applied to maximize influence on static networks can be applied to dynamic networks as well. The network dynamics is measured by the variation of the influence range and absorbed into the learning procedure. Our proposal delicately formulates the effect of network dynamics: the more the influence range varies, the more likely the seeds are to be learned from scratch. Under this assumption, the continuity of the network variation is fully taken advantage of. Experimental results on both synthetic and real-world networks verify the privileges of our proposal against alternative methods.

Maximization

Variation (astronomy)

Scratch

Learning Automata

Dynamics

Social network (sociolinguistics)

10.48550/arxiv.2010.10950

Cite

Citations (0)

MAM: Multiple Attention Mechanism Neural Networks for Cross-Age Face Recognition

Wireless Communications and Mobile Computing (2022)

Xudie Ren Jialve Wang Shenghong Li

Cross-age face recognition problem is of great challenge in practical applications because face features of the same person at different ages contain variant aging features in addition to the invariant identity features. To better extract the age-invariant identity features hiding beneath the age-variant aging features, a deep learning-based approach with multiple attention mechanisms is proposed in this paper. First, we propose the stepped local pooling strategy to improve the SE module. Then by incorporating the residual-attention mechanism, the self-attention mechanism, and the improved channel-attention mechanism to the backbone network, we proposed the Multiple Attention Mechanism Network (MAM-CNN) framework for the cross-age face recognition problem. The proposed framework can focus on essential face regions to highlight identity features and diminish the distractions caused by aging features. Experiments are carried out on two well-known public domain face aging datasets (MORPH and CACD-VS). The results yielded prove that the introduced multiple mechanisms jointly enhance the model performance by 0.96% and 0.52%, respectively, over the state-of-the-art algorithms.

Pooling

Deep Neural Networks

10.1155/2022/8546029

Cite

Citations (4)

The Z Specification Based on ZML

Shenghong Li

The Z specification Standard of ISO provides the fundamentals of developing Z specification tools and inter-operations between different Z specification tools.The standard just established the Z notations which are conformed by different Z specification tools,but did not say how to implement the notations.The Z notations can be exchanged between different Z tools based on ZML that is a XML schema file.Basically,ZML takes advantage of the XML principles of reuse and substitution.The main problems are alignment between ZML notations and Z specification,and reduction of maintenance time of XML document about Z language when a ZML document is evolving.

XML Schema (W3C)

B-Method

Cite

Citations (0)

How to Efficiently Manage Critical Infrastructure Vulnerabilities? Toward Large Code-graph Models

H B Zhang Gaolei Li Shenghong Li Hongfu Liu Shuo Wang

Code (set theory)

10.1145/3689217.3690622

Cite

Citations (0)

Two Approaches on Accelerating Bayesian Two Action Learning Automata

Lecture notes in computer science (2016)

Hao Ge Haiyu Huang LI Yu-lin Shenghong Li Jianhua Li

Learning Automata

10.1007/978-3-319-42297-8_23

Cite

Citations (1)

Mixed geometric loss for bounding box regression in object detection

Journal of Electronic Imaging (2020)

Xudie Ren Fucai Luo Shenghong Li

Predicting bounding box with higher intersection over union (IoU) is one of the most important issues in many computer vision tasks. The ℓn-norm loss and IoU-based loss are two conventional approaches to guide a training process in recent methods. However, the optimization direction of ℓn-norm loss is not exactly the same as maximizing the metric. In addition, IoU-based loss suffers from some inevitable disadvantages due to the direct addition of IoU. According to the shape, size, and position properties, we design a mixed geometric (MG) regression loss to increase the similarity and the overlapping area of two bounding boxes. The shape is described by the cosine similarity of diagonal vectors, the size is measured by the length or width of the boxes, and the location is calculated by the center positions of the boxes. Simulation experiments verify that the proposed MG loss can achieve competitive convergence speed and regression accuracy. By introducing the state-of-the-art models in object detection, experiments are carried out on a well-known benchmark dataset, and the results demonstrate the effectiveness of our method in object detection.

Minimum bounding box

Bounding overwatch

Similarity (geometry)

Benchmark (surveying)

10.1117/1.jei.29.5.053005

Cite

Citations (2)